|
Your single, integrated source for DBAs to manage multi-database, multi-OS
system enterprises from a single source
|
|
|
(PDF; 367K)
 Download
Whitepaper
Surveillance
Technical Whitepaper:
The Evolution of Information (PDF:
2.3M)
|
|
Flexible Agent Architecture
The Surveillance architecture
is client/agent based allowing for n+1 scalability to an unlimited
number of servers.
This allows you to configure your agents to your business requirements,
and you are not bound to some some pre-determined architectural
restriction, such as a centralized console. With Surveillance DB,
you can monitor all of your Oracle, Sybase or SQL Server databases
from either a single
or multiple console configuration through an intuitive graphical
user interface.
Real-Time Monitoring
Surveillance DB is a real-time monitoring tool capable of overseeing
your entire database environment. Surveillance provides a facility to display an extensive
set of predefined windows that provide an immediate global view of database
activity and detailed performance metrics such as session/process activity,
locks, batch contention, file I/O, and much more. Real-time data from
multiple RDBMS can be viewed simultaneously. Data from each window can
be sorted or filtered while most statistics can be graphed over time.
In addition, each window or graph can be configured to refresh at either
the default collection interval or its own refresh interval.
A DBA can view
real-time performance data without writing scripts,
reducing the need for a large, skilled staff to trace problems. Surveillance
deploys an analytical drill-down methodology for quick problem
identification.
For example, Surveillance DB for
Oracle provides a set of pre-defined windows that are used to
immediately display real-time information for a variety of performance
statistics. Highlighting an entity under the Oracle agent displays
a context toolbar at the top of the screen. The buttons on this
toolbar represent the primary windows available for real-time
monitoring. The buttons available depend upon the version of
the Oracle database that is highlighted.
If the first button on the context-specific
toolbar was clicked, a Global window would be opened for the
corresponding database on that server, displaying important statistics
regarding the status of the Oracle instance. Double-clicking
on a line in a window allows you to go deeper or to "drill-down" to
start a time-line graph for that statistic.
The Analyzer
Surveillance DB has an extensive alerting and alarming facility
that comes with a large set of pre-defined collections and rules. To
begin monitoring, the DBA uses these pre-defied rules to easily adjust
default values and collection intervals before turning on the
rules to be monitored.
Predefined and User-defined Collections
The Surveillance DB Collector Agent is responsible
for collecting the performance data from the entity via pre-defined
collections. The data must first be collected by the collector
agent before data can be used by the analyzer or repository
agent for the Event Management and Historical Repository features.
In the past, the end-user was limited to defining perameters
for alerting and adding pre-defined collections to the repository,
and was unable to modify or add new collections. With Surveillance DB,
a facility exists for users to create user-defined collections
that can be used with the Event Management and Historical Repository
features.
Back
to top
Multiple Action Event Handling
The alerter Agents is
an OS process that forwards events to multiple destinations.
Right out of the box, Surveillance DB can be easily set up to
notify
multiple DBAs
by e-mail, pager, HP Openview or Tivoli alerts, pop-up windows,
or to record the error in the Windows Event Log. The DBA can
also execute a fix-it job by executing any operating system executable,
including SQL*Plus and ISQL, or forward events to a central alerter.
Surveillance DB automates
the monitoring process through rules analysis
and event generation. From these events, the system
determines the actions that need to be taken. This process begins
by defining and applying data rules in the Analyzer Agent. When
the agent is started, it loads its rule set, attaches to the
appropriate data collections, and begins a continuous process
of evaluating rules. If a rule has been satisfied (for example,
something noteworthy has occurred) it generates an event that
is sent to the local alerter. The Alerter receives the event,
and determines which, if any, Event Handler is appropriate for
handling the event.
Data Repository
Surveillance DB provides
a mechanism for storing data collections to a repository for
reporting and analysis. This process begins by defining data
stores in the Repository
Agent. When the agent is started, it loads its
list of defined data stores, attaches to the appropriate data
collections, and begins its continuous process of storing data. Data Store
definitions inform the Repository Agent what to
store, how often, and for how long. Data Store definitions must
be specified for the desired server entity and collection.
A Repository Export Program can
be used to export the data obtained in the Repository partitions.
Once the repository data is stored in a comma-delimited ASCII
file, it can be imported to a spreadsheet or database. This facilitates
statistical reporting to analyze the performance data being monitored. The Repository Agent is a process
that stores historical performance or application data from any
pre-defined or user-defined collection. Historical data is retained
in named stores, which have user-specified time intervals and
duration. An interval compression algorithm is used to minimize
disk utilization.
Data from the repository is used
to obtain a historical perspective when the
Event Management detects and alerts the DBA that conditions
are occurring that are negatively impacting performance. Data
from the Repository can be used to trend performance and space
utilization data; or the DBA could use the data to establish
baselines for threshold values in Event Management's rule definitions.
Any ODBC-compliant reporting tool
can be used to query and/or graph the historical data.
Surveillance Reporting
Module
Surveillance allows the transfer of repository data to a central repository
database either by export (push method) or import (pull method). All
supported database systems may be used for the central repository.
The reporting module comes with
a large number of predefined reports which allow out of the box
reporting of the central repository, including scheduling of
reports. The reporting module uses the Crystal Reports engine.
Thus, using the Crystal Reports product, the user can easily
add own reports.
Back to
top
|
